Durham Research Online
Durham Research Online
Deposited in DRO:
18 November 2014
Version of attached le:
Accepted Version
Peer-review status of attached le:
Peer-reviewed
Citation for published item:
Lewald, J. and Kentridge, R. W. and Peters, S. and Tegenthof, M. and Heywood, C. A. and Hausmann, M.
(2013) 'Auditory-visual localization in hemianopia.', Neuropsychology., 27 (5). pp. 573-582.
Further information on publisher's website:
https://doi.org/10.1037/a0033451
Publisher's copyright statement:
c 2013 APA, all rights reserved. This article may not exactly replicate the nal version published in the APA journal.
It is not the copy of record.
Additional information:
Use policy
The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for
personal research or study, educational, or not-for-prot purposes provided that:
• a full bibliographic reference is made to the original source
• a link is made to the metadata record in DRO
• the full-text is not changed in any way
The full-text must not be sold in any format or medium without the formal permission of the copyright holders.
Please consult the full DRO policy for further details.
Durham University Library, Stockton Road, Durham DH1 3LY, United Kingdom
Tel : +44 (0)191 334 3042 | Fax : +44 (0)191 334 2971
http://dro.dur.ac.uk
Lewald, J., Kentridge, R. W., Peters, S., Tegenthof, M., Heywood, C. A. & Hausmann, M. (2013).
Auditory-visual localization in hemianopia. Neuropsychology 27: 573-582..
Auditory-visual localization in hemianopia
Jörg Lewald, Ruhr University Bochum and Leibniz Research Centre for Working
Environment and Human Factors
Robert W. Kentridge, University of Durham
Sören Peters, Ruhr University Bochum
Martin Tegenthoff, Ruhr University Bochum
Charles A. Heywood, University of Durham,
and
Markus Hausmann, University of Durham
Author Note
Jörg Lewald, Department of Cognitive Psychology, Faculty of Psychology, Ruhr University Bochum,
D-44780 Bochum, Germany; and Leibniz Research Centre for Working Environment and Human
Factors, Ardeystr. 67, D-44139 Dortmund, Germany. Sören Peters, Department of Radiology, BGKliniken Bergmannsheil, Ruhr University Bochum, D-44789 Bochum, Germany. Martin Tegenthoff,
Department of Neurology, BG-Kliniken Bergmannsheil, Ruhr University Bochum, D-44789 Bochum,
Germany. Robert W. Kentridge, Charles A. Heywood, Markus Hausmann, Department of Psychology,
University of Durham, Durham DH1 3LE, UK.
Acknowledgement
We thank all participants for their willing cooperation, A. Stöbener and V. Zimmermann for
help with running the experiments, P. Dillmann for preparing the software and parts of the electronic
equipment, and H.-O. Karnath for critical discussion of the results and valuable comments on the
manuscript. This research was supported by the Deutsche Forschungsgemeinschaft (FA 211/24-1).
Correspondence concerning this article should be addressed to Jörg Lewald, Department of
Cognitive Psychology, Faculty of Psychology, Ruhr University Bochum, D-44780 Bochum, Germany.
Email: [email protected]
2
Abstract
Objective: Beyond visual field defects, patients with hemianopia have been suggested
to perceive horizontal visual space in a distorted manner. However, the pattern of these
distortions remained debatable. The aim of this study was to estimate the geometry of the
visual representation of space in hemianopia using an auditory marker.
Method: Patients with pure left or right hemianopia (without neglect) were tested in
tasks requiring them to bring a visual stimulus into spatial alignment with a target sound
(Experiment 1) or vice versa (Experiment 2). In Experiment 1, patients adjusted the location
of a light such that it was displaced towards the anopic side with reference to the physical
sound position. In Experiment 2, patients adjusted the location of a sound such that it was
displaced opposite to the anopic side with reference to the actual position of the visual target.
Results: Both these experiments consistently indicated that hemianopic patients
perceived a sound and a light to be in spatial alignment when the physical position of the light
deviated by several degrees from the sound toward the side of the anopic hemifield, that is, to
the contralesional side.
Conclusions: Given that auditory localization in patients with hemianopia has been
previously shown to be only slightly biased toward the anopic side, the observed distortion of
visual space with reference to auditory space can be explained by assuming that visual
positions were, in absolute terms, perceived as shifted toward the intact side. As a result, HA
patients may perceive visual space as compressed on their ipsilesional (intact), in comparison
with their contralesional (anopic) side.
Keywords: Hemianopia, Visual space perception, Sound localization, Visual cortex,
Multisensory integration
3
Auditory-visual localization in hemianopia
Hemianopia (HA) is a visual field defect characterized by a loss of vision in one
hemifield. The visual defect is caused by unilateral lesions in the cerebral hemisphere
contralateral to the anopic side, either in post-chiasmatic optic tract, lateral geniculate nucleus,
optic radiation, or occipital lobe. HA has a relatively common occurance affecting
approximately 20% of stroke patients and severly affecting their quality of life (Schuett,
Heywood, Kentridge, & Zihl, 2008). It has been suggested that patients with HA perceive
horizontal visual space in a distorted manner. The primary experimental evidence for this
conclusion comes from visual line bisection tasks. With these tasks, HA patients displace the
bisection mark toward the anopic side, which has been interpreted as a bias of visual space
toward the blind field (e.g. Axenfeld, 1894; Liepmann & Kalmus, 1900; Best, 1910, 1917;
Strebel, 1924; Barton & Black, 1998; Hausmann, Waldie, Allison, & Corballis, 2003a). In
approaches employing visual pointing or adjustment tasks, the visual straight-ahead of HA
patients was shown to be displaced toward the anopic side (Zihl & Von Cramon, 1986; Ferber
& Karnath, 1999; Lewald, Peters, Tegenthoff, & Hausmann, 2009a). The exact topography of
the visual space in HA is, however, still a matter of debate. In two studies, the horizontal
angular distance between visual stimuli (Zihl & Von Cramon, 1986) and the horizontal size of
rectangles (Ferber & Karnath, 2001a) was perceived as smaller on the anopic side than on the
intact side, thus suggesting a compression of visual space on the anopic side. Another study
(Doricchi, Onida, & Guariglia, 2002) obtained the opposite result: HA patients estimated size
and distance in the anopic hemifield as being longer than equivalent sizes and distances in
intact hemifield. In each case, the most established result, that visual straight ahead is
displaced to the anopic side in HA, is geometrically compatible only with both a bias of visual
localization to the intact side and a compression and expansion of visual space on the intact
4
and anopic sides, respectively. The reasons for these inconsistencies between studies are still
unclear, but might lie in the fact that previous approaches often did not disentangle a putative
distortion of space from accounts based on spatial attention. In addition, paper/pencil tasks or
tasks with presentation of visual stimuli on a computer screen, as are often employed in this
field of research, might not permit unequivocal interpretations of the topography of visual
space since stimuli are not seen in isolation, but in relation to a visual surround. Finally,
compensatory strategies, utilizing proprioceptive and vestibular cues from the eyes, the head
and the arms (Doricchi et al., 2002), may vary among individuals, in particular depending on
whether patients were tested in the acute or chronic phase of brain injury. Thus, conclusions
about the absolute localization of a visual stimulus in space are difficult to draw from these
previous approaches.
The starting point of the present study was the growing evidence that spatial hearing
performance remains relatively unaffected in HA. Zimmer, Lewald and Karnath (2003) did
not find any significant bias of the auditory median plane in HA patients as measured by
variation of interaural time differences. In a direct comparison, Lewald et al. (2009a) showed
that subjective straight ahead deviates in hemianopia substantially for visual stimuli whereas
it is veridical for auditory stimuli. In a more detailed study (Lewald, Peters, Tegenthoff, &
Hausmann, 2009b) which focussed on the topography of auditory space in HA, there were
statistically significant distortions of auditory space in HA patients which can be interpreted
by both rotation and compression of auditory space toward the anopic side. However, the
mean amplitude of these distortions measured with a task of manual pointing was only 1.5° in
the auditory modality, which is relatively small compared with the visual distortions that have
been recently reported (4-8°; Zihl & von Cramon, 1986; Ferber & Karnath, 1999; Lewald et
al., 2009a, b).
5
In the present study, we thus asked subjects to match the location of a single visual
target with an auditory marker or vice versa in totally dark surroundings. The rationale for our
approach was that accounts in which the relative attentional salience of stimuli with different
spatial locations determines spatial judgements in hemianopia cannot readily be applied to a
task in which there is only a single visual stimulus. For example, an attentional gradient might
affect the way in which patients with hemianopia scan complex visual stimuli. It should not,
however, affect simple alignment judgements containing a single visual stimulus.
Using such simple alignment tasks, the present study sought to establish whether
more general distortions in the visual representation of space accompany deviations in visual
straight ahead. In other words, rather than asking participants to make judgements about the
relative locations of many visual stimuli (as is the case of the visual line bisection task), the
present experiment used an auditory marker to indicate the location of a single visual
stimulus. From such judgements in HA patients and normal observers, we aimed to estimate
the potential distortions in the representation of visual space accompanied by HA.
Method
Subjects
Results from ten patients with brain lesions were included in this study. All patients
had received the diagnosis of persistent homonymous hemianopia (HA) confined to one
hemifield, as confirmed by visual perimetry tests (see below). HA was left-sided (LHA) in
seven patients (LHA1-LHA7) and right-sided (RHA) in three patients (RHA1-RHA3). Details
on age, sex, visual field defects, and lesion sites are reported in Table 1 and Suppl. Table 1.
All patients were congenitally right-handed, as assessed by a German adaptation of Coren’s
(1993) inventory (Siefer, Ehrenstein, Arnold-Schulz-Gahmen, Sökeland, & Luttmann, 2003),
with a criterion of an individual score of ≥ 2 (range from -4 to 4) in the handedness section of
6
this questionnaire. However, hemiparesis prevented two patients (LHA5, RHA2) from use of
their contralesional hand, and one other patient showed mild (LHA7) impairment with use of
the contralesional hand due to hemiparesis. In addition to these ten participants with HA, two
further patients (one LHA and one RHA patient) were also tested, but were excluded from the
study since they were unable to adequately perform the acoustic pointing task, in particular
when visual targets were presented on the anopic side. For these two subjects, linear
regression analyses of the pointing responses to targets on the anopic side resulted in rather
poor coefficients of determination (R2 of .06 and .17), which were more than four standard
deviations below the R2 values of the ten subjects included (range from .65 to .92, mean .78,
SD .10; see 2.).
All HA patients had circumscribed brain lesions as a result of ischemic stroke or
hemorrhage, demonstrated by magnetic resonance imaging (MRI) or computed tomography
(CT). In all patients, lesions were unilateral (i.e., on the side contralateral to the anopic
hemifield), with the exception of patient RHA3 who showed some minor involvement of
right-hemispheric regions in addition to the predominant left-hemispheric lesion. Lesion sites
of all patients are summarized in detail in Suppl. Table 1 and Suppl. Figure 1.
To test whether HA patients suffer from spatial neglect a neglect-test battery (Ferber
& Karnath, 2001a) was applied, which consisted of the following tests: (a) Letter Cancellation
task (Weintraub & Mesulam, 1985), which requires the patient to cancel 60 target letters ‘A’
distributed amid distractors on a horizontally oriented standard page (DIN A4). Responses
were coded and the Center of Cancellation (CoC) index was measured using the software
(www.mricro.com/cancel/) by Rorden and Karnath (2010). Patients are classified as suffering
from spatial neglect when they show a CoC index > .09 (Rorden & Karnath, 2010). (b) Bells
Test (Gauthier, Dehaut, & Joanette, 1998), which requires the patient to identify 35 bell
symbols distributed on a horizontally oriented standard page with 40 distractor symbols.
7
Responses were analysed by calculating the CoC index (Rorden & Karnath, 2010) as with the
Letter Cancellation Task, using the same cutoff threshold (> .09) for the patient's
classification as suffering from spatial neglect. (c) Baking Tray Test (Tham & Tegnér, 1996),
which requires the patient to place 16 identical items as evenly as possible on a blank standard
page (8 on the left, and 8 on the right side). Any distribution more skewed than seven items
on the left side and nine items on the right side are considered as a sign of spatial neglect. (d)
Copying task (Ferber & Karnath, 2001a; Johannsen & Karnath, 2004), in which patients are
asked to copy a complex multi-object scene consisting of four figures on a standard page (two
on the left, and two on the right side). Omission of one left sided feature of each figure is
scored as 1, and omission of each whole figure is scored as 2, resulting in a maximum score
of 8. A score higher than 1 (i.e. > 12.5% omissions) is considered as a sign of spatial neglect.
None of the HA patients exceeded the limit values in at least two of these four tests, which
has been regarded as the criterion for presence of spatial neglect (Karnath, Himmelbach, &
Rorden, 2002).
In addition, we applied a line-bisection task which comprised 17 horizontal black lines
of 1 mm width on a horizontally-oriented white standard page. The lines ranged from 100 to
260 mm long, in steps of 20 mm. The mean length was 183.5 mm. Patients were asked to
bisect all lines into two parts of equal length by marking the subjective midpoint of each line
with a fine pencil (for details, see e.g. Hausmann et al., 2003a; Hausmann, Corballis, & Fabri,
2003b). The majority of neglect patients shows a large bisection bias towards the right,
although about 30% of patients with acute neglect do not show any significant bias in linebisection tasks (Ferber & Karnath, 2001b). Here, HA patients showed a significant mean
bisection bias of 4.77% (SE 1.67, range from -2.15% to 11.32%), t(9) = 2.85, p = .019, toward
the side of the anopic hemifield (LHA: mean leftward bias -6.58%, SE 1.96; RHA: mean
8
rightward bias .53%, SE 1.47). This conforms with previous findings of a contralesional bias
in patients with HA (e.g., Barton & Black, 1998; Hausmann et al., 2003a,b).
Prior to experimentation, the presence of homonymous HA was confirmed by visual
static perimetry tests in all patients included in this study. In addition, after completion of the
experiments the azimuthal dimensions of the visual field, and in particular the position of the
binocular vertical visual field border (Table 1) was measured in more detail using visual
stimulation by the experimental apparatus, as was already described in preceeding studing
(Lewald et al., 2009b; Lewald, Tegenthoff, Peters, & Hausmann, 2012). For this purpose,
white light flashes (duration 50 ms), delivered by light-emitting diodes (LEDs; see below),
were presented in total darkness at random locations in the azimuthal plane over a range
from -90° on the left to 90° on the right, in steps of 2°. Patients were instructed to fixate on a
central red light emitting diode, that was permanent on, and to press a response button as soon
as they perceived a white light flash. Stimuli were presented with a randomly varied time
interval between 1 s and 3 s (steps of .5 s) after the patients' response. In one block, each
stimulus position was presented three times, resulting in 273 trials. Data of four identical
blocks (2 blocks conducted on one day and 2 blocks on a separate day) were pooled. For
computation of the visual field border, the number of correct responses was plotted as a
function of stimulus azimuth (θ) within the range of -46° on the left to 46° on the right, and
fitted to the sigmoid equation:
f = 100 / (1 + e -k(θ - VFB))
where f is the frequency of responses, given as percentage; VFB (visual field border)
is that θ where f is 50%; k is the slope of the function at 50%; e the base of the natural
logarithm (Lewald et al., 2009b, 2012). The mean coefficient of determination (R2) of the fit
was .93 (range from .67 to 1.00; all p < .0001), indicating a sharp boundary of the visual field
for all patients. Patients detected the vast majority of stimuli in the intact hemifield (mean
9
85.7%, SE 3.6) and only few in the anopic hemifield (7.7%, SE 2.2; measured over the range
from 90° to 2° eccentricity on the respective side). Across all patients, the VFB was only
slightly shifted toward the side of the anopic field (mean 3.38°, SE 1.28). One of the patients
(LHA7) showed incomplete left HA, with a small peripheral area of vision lying to the left of
the anopic field.
Ten healthy right-handed subjects (4 females and 6 males), ranging in age from 39 to
66 years (mean 49.9 years, SE 3.4), participated in the study as normal controls. Each control
subject was matched with one of the 10 HA patients for sex and age (±3 years).
All subjects, HA patients and normal controls, were tested for general hearing loss.
For this purpose, white-noise bursts with a duration of 1 s were presented monaurally via
headphones (K271, AKG Acoustics, Vienna, Austria) at various sound-pressure levels (SPLs,
range 10-80 dB re 20 µPa, steps of 10 dB; onset/offset time 50 ms), and subjects pressed a
button as soon as they heard a sound. A two-factor ANOVA with Ear (ipsilateral,
contralateral) as within-subject factor and Group (LHA patients, RHA patients, controls) as
three-level between-subjects factor revealed neither a main effect nor interaction, all F ≤ 2.77,
p > .09. Most importantly, HA patients did not show any superiority of the ear on the side of
the intact (contralateral) or the anopic (ipsilateral) hemifield, t(9) = .00, p = 1.00.
A subsequent hearing test was focussed on the symmetry in loudness perception of the
subjects' left and right ears, which is more relevant to the experiments conducted here than
thresholds. For this purpose, incoherent white-noise signals (preventing binaural fusion) were
presented binaurally via headphones (as above). Interaural SPL (average root mean square)
differences for these stimuli were varied between trials following a quasi-random order over a
range from 20 dB (higher SPL at the left ear) to 20 dB (higher SPL at the right ear), in steps
of 4 dB (duration 1 s; onset/offset time 50 ms; mean SPL 70 dB). Subjects were instructed to
make a two-alternative forced choice as to which of the two sounds was louder, the one on the
10
left or the one on the right. The test was composed of 110 trials (10 presentations of each level
difference) and lasted about 5 min. The point of subjective equality measured in HA patients
(LHA: mean .05 dB, SE 1.33; RHA: mean -1.12 dB, SE 1.60) and controls (mean -0.73 dB,
SE .58) did not differ, F(2,17) = .45, p = .64, nor was there any bias to the side of the intact or
the anopic hemifield in HA patients, t(9) = .37, p = .72. Taken together, with respect to these
basic auditory tests, the HA patients' auditory performance of both ears was symmetrical and
normal.
This study conformed to the Code of Ethics of the World Medical Association
(Declaration of Helsinki), printed in the British Medical Journal (18 July 1964). All subjects
gave their informed consent to participate in the study, which was approved by the Ethical
Committee of the Medical Faculty of the Ruhr University Bochum.
Apparatus
The experiments took place in a sound-proof and anechoic room (5.4 × 4.4 × 2.1 m3),
which was insulated by 40 cm (height) × 40 cm (depth) × 15 cm (width at base) fiberglass
wedges on each of the six sides. A suspended mat of steel wires served as the floor. The
ambient background noise level was below 20 dB(A) SPL. All experiments were conducted in
total darkness.
The acoustic stimulus was band-pass-filtered noise (lower cut-off frequency .8 kHz;
upper cut-off frequency 3 kHz) with a maximum duration of 12 s (rise/fall time 100 ms).
Sound stimuli were generated digitally and converted to analog form via a computercontrolled external soundcard (Sound Blaster Audigy 2 NX, Creative Labs, Singapore) at a
sampling rate of 96 kHz. Sound stimuli were delivered via a semicircular loudspeaker system,
with an SPL of 75 dB. The subject sat on a comfortable chair. In front of the subject at a
constant distance of 1.5 m from the centre of the head, 91 broad-band loudspeakers (5 × 9
11
cm2, Visaton SC 5.9, Visaton, Haan, Germany) were mounted in the subject's horizontal
plane. The azimuth of the loudspeakers ranged from -90° (left) to 90° (right), in steps of 2°,
with the centre loudspeaker at 0°. For visual stimulation at corresponding azimuthal positions,
at the lower edge of the chassis of each loudspeaker a white LED was mounted in a central
position. The LED (diameter 10 mm; luminance about 100 cd/m2) was mounted in a small
housing impermeable to light, with a central circular aperture of 2 mm diameter immediately
in front of the LED.
Procedure for Experiment 1: Visual pointing to acoustic targets
The subject's head was fixed by a custom-made framework with stabilizing rests for
the chin, forehead, and occiput (see Lewald, 1997). In Experiment 1, subjects had to bring a
visual stimulus into spatial alignment with a target sound. This task is a modification of the
method originally described in Lewald and Ehrenstein (1998). In each trial, a stationary target
sound was presented. Acoustic stimuli were presented from 21 loudspeaker positions: straight
ahead of the subject (0°), 10 positions on the left and 10 positions on the right with constant
angular separation of 4°, thus covering an angular range from 40° to the left to 40° to the
right. Each trial began with the onset of the sound stimulus at one of the 21 positions. The
stimulus position changed in a quasi-random order between trials. At the moment of sound
onset, a continuous light stimulus, delivered from an LED, was presented simultaneously at
one out of nine locations (from -24° to 24° azimuth, with angular separation of 6°). The initial
position of the light was varied following a quasi-random order. The subject controlled the
azimuthal position of the light (over a total range of 180°, in steps of 2°) by adjusting the
knob of a potentiometer. The potentiometer was mounted in a small case, so that the subject
held it in one hand while turning the knob with the other hand (see Fig. 2 in Lewald et al.,
2009a). Subjects were instructed to direct their gaze to the light and, while maintaining
fixation on it, to adjust its position (by turning the knob of the potentiometer) toward the
12
position of the sound until the locations of both these stimuli were perceived to be in exact
alignment. HA patients were explicitly encouraged to search for the visual stimulus by eye
movements, as it could start in their anopic field. The subjects were instructed to press a
button (mounted beside the potentiometer knob on the case) as soon as the adjustment was
completed. At the moment the key was pressed, both the light and the sound disappeared and
the final position of the visual stimulus was recorded. Each of the 21 loudspeaker positions
was presented in combination with each of the nine starting positions of the LED, thus
resulting in a total number of 189 trials plus repetitions. Two seconds after key pressing, the
next trial began. The sound and light stimuli had a maximum duration of 12 s. After about 40
practice trials, all subjects were able to perform the task within about 5-8 s. In cases in which
the key was not pressed before the stimulus ended automatically (that is, within 12 s), the trial
was repeated at the end of the experiment. Each experiment comprised 168 trials plus
repetitions (eight presentations of each stimulus position). The timing of the stimuli and the
recording of the subject’s responses were controlled by custom-written software.
Procedure for Experiment 2: Acoustic pointing to visual targets
In Experiment 2, subjects were asked to bring an acoustic stimulus into spatial
alignment with a visual target. The task was conducted analog to that described for
Experiment 1, with the only difference that visual and auditory stimuli were interchanged.
Stationary visual target stimuli were presented from 21 LED positions: straight ahead of the
subject (0°), 10 positions on the left and 10 positions on the right with constant angular
separation of 4° (from -40 to 40° azimuth). The light and sound stimuli had a maximum
duration of 12 s. The position of the target light changed in quasi-random order between trials.
In each trial, at the moment of light onset a continuous sound stimulus (as described above)
was simultaneously delivered from a loudspeaker at one out of nine locations (from -24° to
24° azimuth, with angular separation of 6°). The initial position of the sound was varied
13
following a quasi-random order. The subject controlled the azimuthal position of the sound
(over a total range of 180°, in steps of 2°) by adjusting the knob of the potentiometer in an
identical manner as described above for adjustment of light stimuli (see 2.3.). Subjects were
instructed to direct their gaze to the light and, while maintaining fixation on it, to adjust the
sound position (by turning the knob of the potentiometer) toward the position of the light until
the locations of both these stimuli were perceived to be in exact alignment. The subjects
pressed the button as soon as the adjustment was completed. HA patients were able to
perform this task with similar ease as that in Experiment 1. At the moment the key was
pressed, both the light and the sound disappeared and the final position of the auditory
stimulus was recorded automatically. Each of the 21 LED positions was presented in
combination with each of the nine starting positions of the sound, thus resulting in a total
number of 189 trials plus repetitions. All other parameters and conditions were identical to
those in Experiment 1.
Data analysis
For analysis of the data obtained in Experiments 1-2, the subject's individual pointing
responses were determined as a function of target position, and were fitted to a regression
line. Data obtained for stimuli presented in the left and right hemispaces were analysed
separately. Responses were normalized such that positive angles indicate pointing toward the
hemispace within which the stimulus was presented, and negative angles indicate pointing
responses to the opposite hemispace (cf., e.g., Fig. 1). Three parameters, derived from the fit,
were used to describe different aspects of the subject's individual performance. (1) The
y-intercept of the regression line was taken as a measure of the subject's constant error in
pointing to either side. (2) The slope of the regression line (a) was taken as a measure of the
subject's general tendency to underestimate (a < 1) or overestimate (a > 1) the distances
14
between target positions. (3) The coefficient of determination (R2) was taken as a measure of
the subject's precision in pointing.
For statistical comparisons, results of LHA and RHA patients, obtained in
Experiments 1-2, were normalized and pooled. For this purpose, data were classified
according to whether they had been obtained within the hemispace of the patient's anopic or
intact field. As already mentioned, each HA patient was assigned to a healthy control subject
matched for age and sex. Each data set of the control subject was treated in exactly the same
way as the data of the related patient. That is, as data obtained for the left (right) hemispace in
LHA patients and data obtained for the right (left) hemispace in RHA patients were pooled,
the data obtained for the left (right) hemispace of the matched LHA controls and the data
obtained for the right (left) hemispace of matched RHA controls were also pooled. This was
mainly done in order to account for effects of handedness on analyses of the normalized data.
Furthermore, y-intercepts resulting from analyses were normalized such that positive
values indicate a bias in pointing toward the anopic side and negative values a bias toward the
intact side. Finally, to adequately compare the y-intercepts obtained in Experiment 1 with
those of Experiment 2, these values were normalized such that in both experiments positive
values indicate final deviations of the visual stimulus from the auditory stimulus toward the
anopic side, irrespective of whether the visual stimulus was aligned with the auditory target
(Experiment 1) or the auditory stimulus was aligned with the visual target (Experiment 2).
At the first stage of statistical analysis, multi-factor analyses of variance (ANOVAs)
were conducted in order to compare performances of HA patients and controls. In subsequent
stages of analysis, one-factor ANOVAs were used to reveal differences between the
performances measured in the intact hemispace and in the anopic hemispace of HA patients.
For all computations, F-statistics were based on ε-corrected degrees of freedom
15
(Greenhouse-Geisser correction). Bonferroni-corrected p-values were used for multiple
comparisons.
Results
Although HA patients had some difficulties in performing these tasks, linear
regression of the pointing responses as a function of target azimuth was significant (p <
.0001) for all participants, both in Experiment 1 (HA patients: mean R2 = .87, range from .48
to .96; controls: mean R2 = .93, range from .82 to .97) and in Experiment 2 (HA patients:
mean R2 = .80, range from .65 to .94; controls: mean R2 = .87, range from .71 to .95).
A 2 × 2 × 2 mixed ANOVA with Task [visual pointing, acoustic pointing] and
Hemispace [anopic, intact] as within-subject factors and Group [HA, controls] as
between-subjects factor was conducted for the coefficient of determination (R2) of the linear
regression. The ANOVA revealed a main effect of Task, F(1,18) = 12.32, p = .003, ηp2 = .41,
indicating a generally higher precision with light pointing than with acoustic pointing (Fig. 1,
2). No further main effect or interaction was significant (all F ≤ 3.71).
For the normalized y-intercept resulting from the linear regression, an analogous
ANOVA was conducted. The ANOVA revealed a main effect of the factor Group, indicating
the general bias in adjustments with visual stimuli shifted, with reference to auditory stimuli,
to the side of the anopic hemifield, F(1,18) = 12.39, p = .002, ηp2 = .41. Furthermore, a main
effect of Hemispace, F(1,18) = 6.19, p = .023, ηp2 = .26, was in alignment with the general
asymmetry in displacements. Finally, a three-way interaction of Task × Hemispace × Group,
F(1,18) = 5.08, p = .037, ηp2 = .22, was found. Taken together, the findings of this ANOVA
confirmed the obvious influence of HA on cross-modal constant error, as obtained
concordantly in both tasks: Stimulus pairs were adjusted such that visual stimuli were shifted
toward the anopic side with reference to the auditory stimuli. As confirmed by the three-way
16
interaction, this bias was stronger in the patients' anopic hemispace than on the intact side,
and the bilateral asymmetry was more prominent with the light-pointing task than with
acoustic pointing (Figs. 1-3).
Finally, an ANOVA, computed for the slope of the regression line, revealed a main
effect of the factor Task, F(1,18) = 31.33, p < .0001, ηp2 = .64, confirming that in the acoustic
pointing task (Experiment 2) lateral target positions were increasingly underestimated with
increasing eccentricity, whereas in the light pointing task (Experiment 1) the lateral target
positions were increasingly overestimated with increasing eccentricity (cf. Figs. 1, 2). In
addition, a Task × Hemispace × Group threefold interaction, F(1,18) = 13.74, p = .002, ηp2 =
.43, indicated a differential influence of HA on the slope in both tasks: The slope obtained
with light pointing was decreased in the patients' anopic hemifield and increased in the intact
hemifield with reference to healthy controls, whereas the opposite pattern was found with
acoustic pointing (Fig. 3 C, D).
If the ANOVAs for all three dependent variables were restricted to HA patients with
left hemispheric lesions (together with corresponding control subjects), the significance of the
results (not shown here) remained essentially unchanged (the 3-way interaction for the yintercept dependent variable only approached significance, F(1,12) = 4.43, p = .057).
If the three patients with lesions involving parietal areas and their respective controls
were excluded and the analyses were restricted to patients with temporal and/or occipital
lesions, all main effects and interactions with Group as a factor remained the same when
analysing the slope and R2. The only difference occurred for the y-intercept, for which the
three-way interaction of Task × Hemispace × Group only approached significance, F(1,12) =
3.50, p = .086, ηp2 = .23.
17
These results, and in particular the three-way interaction are complicated by the fact
that the slope for the auditory pointing task is in terms of degrees of auditory pointing location
per degree of change in visual location whilst the slopes for the visual pointing task are in the
inverse units. One means of clarifying the results for slope is to recast the analysis so that, for
both tasks, the dependent variable slope is always the change in position of the auditory
stimulus obtained for a given change in location of the visual stimulus (i.e., as in Fig. 3C and
3D). When we did this, the analysis revealed a significant effect of Task, F(1,18) = 15.64, p <
.001, and an interaction between Group and Hemispace, F(1,18) = 15.08, p < .001. The threeway interaction (which in the original analysis merely reflected the fact that the auditory
location and visual location axes were interchanged in the two tasks) disappeared. These
results clarified the finding from the untransformed data insofar as visual space relative to the
auditory representation of space was compressed more strongly on the intact, than on the
anopic, side in HA patients, while normal controls performed essentially symmetrically.
Subsequent analyses for the group of HA patients were conducted for the three
parameters resulting from the linear regression, using one-factor ANOVAs with Hemispace as
factor. The analysis for the coefficient of determination did not provide significant differences
between hemispaces, thus suggesting equal precision in pointing in both hemispaces
(Experiment 1: F(1,9) = 2.20, p = .17, ηp2 = .20; Experiment 2: F(1,9) = 2.70, p = .14, ηp2 =
.23). However, an analogous ANOVA revealed a significant difference in the position of the
y-intercept between hemispaces for Experiment 1, F(1,9) = 7.29, p = .024, ηp2 = .45: The bias
of visual pointing, with reference to auditory targets, toward the anopic side was stronger
within anopic hemispace (mean 6.72°, SE 1.15) than within intact hemispace (mean 2.96°, SE
.98; Fig. 3). In Experiment 2, the bias of acoustic pointing was almost equal in anopic (mean
4.39°, SE .83) and intact hemispaces (mean 3.83°, SE .70), F(1,9) = .64, p = .44, ηp2 = .07
(Fig. 3). Finally, an analogous one-factor ANOVA with Hemispace as factor indicated a
18
significantly steeper slope of the regression line in anopic (mean .86, SE .06), than in intact,
hemispace (mean .77; SE .05) in Experiment 2, F(1,9) = 14.25, p = .004, ηp2 = .61, thus
suggesting that the effects of HA partially counteracted the normally observed pattern of
increasing underestimation with increasing target eccentricity (see above). For Experiment 1,
this approached significance, F(1,9) = 3.81, p = .08, ηp2 = .30, suggested that the normally
observed increase in overestimation with increasing target eccentricity was partially reduced
in anopic hemispace (see above).
It is important to note that these results describe divergences between visual and
auditory spatial representations, but not deviations of perceptual from physical spatial
coordinates. As a consequence, these findings, if considered in isolation, did not allow any
conclusions of whether HA patients showed perceptual anomalies in audition or vision, or in
both of these modalities. To clarify this problem, data obtained in one modality are needed in
addition. In our preceeding study (Lewald et al., 2009b), all HA patients included here had
been tested for auditory localization by using a task of hand pointing to acoustic targets. In
that study, these ten individuals showed a constant error in pointing toward the anopic side,
that was, however, relatively small in amplitude (anopic side: mean 2.63°, SE 1.54; intact
side:mean .42°, SE 1.81). A statistical comparison of the intermodal divergence between
visual and auditory locations (mean normalized y-intercepts from Experiments 1 and 2; Fig
3A, B) and the unimodal auditory y-intercepts taken from Lewald et al. (2009) for both
hemispaces revealed a significantly larger bias in the present study (mean 4.47°, SE .67) than
in the preceeding study (mean 1.52°, SE .50), t(9) = 4.00, p = .003. Also, the slope obtained
by Lewald et al. (2009b) for the regression line of hand pointing responses as a function of
auditory target position (mean .94, SE .05) was significantly flatter than the mean slope
obtained here after conversion of data such that visual positions were always plotted as a
function of sound position (mean 1.18, SE .06), t(9) = 3.93, p = .003. Thus, given the previous
19
unimodal auditory results from the same HA patients (Lewald et al., 2009b), the already
known distortion of their auditory space significantly differed from the intermodal distortion
found here.
Discussion
These results demonstrated a significant distortion of visual space with reference to
auditory space in patients with pure HA. First, HA patients generally perceived visual
locations to be displaced toward their intact hemifield. Secondly, HA patients perceived
visual space relative to the auditory representation of space as more compressed in their
intact, than in their anopic, hemifield.
The interpretation of these findings is complicated by the fact that HA patients might
potentially exhibit distortions not only in their visual representations of space (Zihl & von
Cramon, 1986; Ferber & Karnath, 1999) but also in the auditory modality (Lewald et al.,
2009a, b), although these latter anomalies seemed to be relatively slight. Nevertheless,
unimodal auditory distortions of space perception cannot explain the intermodal divergences
found here. The same patients, who were participants in the current study have also been
tested in hand pointing to auditory targets (Lewald et al., 2009b). The extent of the unimodal
distortion found in that previous study cannot account for the intermodal distortion found
here. Lewald et al. (2009b) reported constant errors that were about half the values obtained
here for intermodal bias. This suggests that the results obtained in the current study had their
origin primarily in the anomalies of visual spatial perception, rather than in the auditory
domain. Lewald et al. (2009a) found the auditory straight ahead of HA patients not to differ
from that of normal controls. It may therefore be reasonable to conclude that constant errors
in auditory localization, although statistically significant, are small in magnitude compared
with the distortions of visual space in HA. If one assumes that auditory space perception is
only subject to relatively small errors, the pointing bias obtained in both experiments is
20
consistent with a distortion of visual space in which visual positions were mislocalized toward
the unimpaired hemifield (Fig. 3A, B), suggesting visual space was perceived as compressed
on the intact, compared with the anopic, hemifield (Fig. 3C, D).
This conclusion is in line with previous findings on the position of the visual straight
ahead in pure HA. Several studies consistently demonstrated a bias of the visual straight
ahead toward the anopic side (Zihl & von Cramon, 1986; Ferber & Karnath, 1999; Lewald et
al., 2009a). That is, if a visual stimulus is physically located straight ahead, it will be
mislocalized toward the intact side. The amplitude of the shift in visual straight ahead was
reported to be about 4-8°, which is compatible with the average bias of 4.47° obtained here
given a substraction of the mean unimodal auditory bias of 1.52° measured by Lewald et al.
(2009b). These earlier results are, however, based on estimates of visual locations with
reference to the subjective coordinates of the body. They could be confounded by a
proprioceptive bias that might occur in addition to visual anomalies with HA (Lewald et al.,
2009a). Such proprioceptive factors cannot account for the results obtained in the current
study.
Unlike the previous findings on visual straight ahead, which were restricted to one
central point in space only, the present results provide information on a broader topography of
visual space in HA patients. Taken together, our data indicate that the rotation of the visual
space along the horizontal axis results in a perceptual expansion in the contralesional (blind)
hemifield and a compression in the ipsilesional (intact) hemifield. There is a conflict in the
literature about the nature of visual space distortion in HA. Our conclusion is consistent with
findings of Doricchi et al. (2002), showing that HA patients estimated lengths and distances in
the contralesional space as being larger than their equivalents in the ipsilesional space. Our
results are not consistent with studies showing that the horizontal angular distances between
visual stimuli (Zihl & Von Cramon, 1986) or sizes of rectangles along the horizontal axis
21
(Ferber & Karnath, 2001a) were perceived as smaller on the anopic hemifield than on the
intact side. The reasons for this inconsistency between previous studies are not entirely clear
(see discussion in Ferber & Karnath, 2001a; Doricchi et al., 2002).
From a methodological point of view, it is important to emphasize that in our
experiments patients fixated a single light spot in total darkness. In the studies of Zihl & Von
Cramon (1986), Ferber & Karnath (2001a), and Doricchi et al. (2002) patients had to judge
distances between simultaneously presented visual stimuli or the size of visual objects in
space. In the present study, subjects were presented with single punctiform visual stimulus in
otherwise empty space. Their pointing responses (whether with a visual pointer or with a
visual target) are more likely to reflect absolute judgements of localization than judgements of
the relatiave locations of pairs of points. Relative spatial judgement may enage processes
above and beyond those required to make a simple localization. Our results may therefore
provide a more direct estimate of space distortion in HA.
Several methodological issues have to be considered given our use of two
complementary tasks, The analyses of the coefficient of determination (R2) showed that the
acoustic-pointing responses were more variable than light-pointing responses. This may
reflect greater uncertainty in the localization of the acoustic pointer compared with the light
pointer and the fact that subjects were, doubtlessly, more familiar with pointing to objects
visually in everyday life. Critically, however, there were no significant differences in
variability between groups, suggesting that these task differences did not have a differentially
strong effect on HA patients. These differences in task difficulty may have contributed to the
task-related differences in slope as found in both normal controls and HA patients. While light
pointing was nearly veridical, with acoustic pointing subjects generally underestimated the
eccentricity of the target, resulting in flatter slope of regression lines (Fig. 1, 2). Most likely,
the acoustic marker was not moved far enough due to the greater uncertainty in this task. In
22
the acoustic pointing task, the stationary visual target, which may have been in a location
initially invisible to HA patients, had to be localized at the start of a trial. This imposed a
visual search demand that was not present in the light pointing task. There was, however, no
time pressure to respond and the opportunity to repeat the trial if the target was not perceived
in time. The results (see Fig. 2) suggest that HA patients did not have specific problems in
acoustic pointing (Experiment 2), compared with visual pointing (Experiment 1).
There is another potential problem with using two modalities simultaneously. In the
so-called ventriloquism effect the location of an auditory stimulus is captured by a
simultaneously presented visual stimulus. This effects is, however, unlikely to have affected
results in the current study because it is critically dependent on synchronized transient or
modulated signals in the auditory and visual modalities (see, e.g., Lewald, Ehrenstein, &
Guski, 2001). In our experiment, visual and auditory stimuli were present continously and so
no such synchronized transients occurred.
The patients' perceptual anomalies found here provide direct evidence for substantial
differences in the distortions of sensory space between auditory and visual modalities. There
is already a large body of evidence showing significant differences between uni-modal and
cross-modal processing in brain-damaged patients, including those with visual field defects.
In fact, although HA patients may exhibit distortions in their visual representations of space as
well as in the auditory modality, their cross-modal abilities might be preserved (e.g.,
Bolognini, Rasi, Coccia, & Làdavas, 2005; Leo, Bolognini, Passamonti, Stein, & Làdavas,
2008; Passamonti, Frissen, & Làdavas, 2009). However, had the distortions of auditory and
visual space been essentially similar, then in our bimodal approach they would have cancelled
each other out. Pointing responses would appear to be veridical as the coordinates in the
distorted auditory and visual spaces would nevertheless be congruent with one another. Our
finding that these space distortions cannot be similar matches the direct comparison by
23
Lewald et al. (2009a) who showed a considerable divergence between the subjective straightahead directions of HA patients in the visual and auditory modalities. Based on the earlier
literature on distortions of visual space in HA (see above), Lewald et al. (2009b) originally
assumed that processes of cross-modal spatial adaptation induced a visual miscalibration of
auditory space, such that its coordinates would be slightly shifted toward the point of
alignment with the distorted visual coordinates. Although the present experiments were not
intended to test this hypothesis, the results shed doubts on its validity. It seems reasonable to
conclude that the auditory space of patients with pure HA remained largely unaffected by the
consistent auditory-visual disparity. If HA patients retain an undistorted representation of
auditory space, then it should be possible to exploit the auditory system in rehabilitation of
visual field disorders, even when very severe impairment of visual abilities limits the
effectiveness of purely visual approaches (Lewald et al., 2012).
24
References
Axenfeld, D. (1894). Eine einfache Methode Hemianopsie zu constatiren.
Neurologisches Centralblatt, 13, 437-438.
Barton, J.J., & Black, S.E. (1998). Line bisection in hemianopia. Journal of
Neurology, Neurosurgery, and Psychiatry, 64, 660-662.
Best, F. (1910). Bemerkungen zur Hemianopsie. Graefes Archiv für
Ophthalmologie, 74, 400-410.
Best, F. (1917). Hemianopsie und Seelenblindheit bei Hirnverletzungen.
Graefes Archiv für Ophthalmologie, 93, 49-150.
Bolognini, N., Rasi, F., Coccia, M., & Làdavas, E. (2005). Visual search
improvement in hemianopic patients after audio-visual stimulation. Brain, 128, 26302642.
Coren, S. (1993). The lateral preference inventory for measurement of
handedness, footedness, eyedness, and earedness: norms for young adult. Bulletin of
the Psychonomic Society, 31, 1-3.
Damasio, H., & Damasio, A.R. (1989). Lesion Analysis in Neuropsychology.
New York: Oxford University Press.
Doricchi, F., Onida, A., & Guariglia, P. (2002). Horizontal space
misrepresentation in unilateral brain damage. II. Eye–head centered modulation of
visual misrepresentation in hemianopia without neglect. Neuropsychologia, 40, 11181128.
25
Ferber, S., & Karnath, H.-O. (1999). Parietal and occipital lobe contributions
to perception of straight ahead orientation. Journal of Neurology, Neurosurgery, and
Psychiatry, 67, 572-578.
Ferber, S., & Karnath, H.-O. (2001a). Size perception in hemianopia and
neglect. Brain, 124, 527-536.
Ferber, S., & Karnath, H.-O. (2001b). How to assess spatial neglect - line
bisection or cancellation tasks? Journal of Clinical and Experimental
Neuropsychology, 23, 599-607
Gauthier, L., Dehaut, F., & Joanette, Y. (1998). The bells test: a quantitative
and qualitative test for visual neglect. International Journal of Clinical
Neuropsychology, 11, 49-54.
Hausmann, M., Waldie, K.E., Allison, S.D., & Corballis, M.C. (2003a). Line
bisection following hemispherectomy. Neuropsychologia, 41, 1523-1530.
Hausmann, M., Corballis, M.C., & Fabri, M. (2003b). Line bisection in the
split brain. Neuropsychology, 17, 602-609.
Johannsen, L., & Karnath, H.-O. (2004). How efficient is a simple copying
task to diagnose spatial neglect in its chronic phase. Journal of Clinical and
Experimental Neuropsychology, 26, 251-256.
Karnath, H.-O., Himmelbach, M., & Rorden, C. (2002). The subcortical
anatomy of human spatial neglect: putamen, caudate nucleus and pulvinar. Brain, 125,
350-360.
Leo, F., Bolognini, N., Passamonti, C., Stein, B.E., & Làdavas E. (2008).
Cross-modal localization in hemianopia: new insights on multisensory integration.
Brain, 131. 855-865.
26
Lewald, J. (1997). Eye-position effects in directional hearing. Behavioural
Brain Research, 87, 35-48.
Lewald, J., & Ehrenstein, W.H. (1998). Auditory-visual spatial integration: A
new psychophysical approach using laser pointing to acoustic targets. Journal of the
Acoustical Society of America, 104, 1586-1597.
Lewald, J., Ehrenstein, W.H., & Guski, R. (2001). Spatio-temporal constraints
for auditory-visual integration. Behavioural Brain Research, 121, 69-79.
Lewald, J., Peters, S., Tegenthoff, M., & Hausmann, M. (2009a). Dissociation
of auditory and visual straight ahead in hemianopia. Brain Research, 1287, 111-117.
Lewald, J., Peters, S., Tegenthoff, M., & Hausmann, M. (2009b). Distortion of
auditory space in hemianopia. European Journal of Neuroscience, 30, 1401-1411.
Lewald, J., Tegenthoff, M., Peters, S., & Hausmann, M. (2012). Passive
auditory stimulation improves vision in hemianopia. PLoS ONE, 7, e31603.
Liepmann, H., & Kalmus, E. (1900). Über eine Augenmassstörung bei
Hemianopikern. Berliner Klinische Wochenschrift, 37, 838-842.
Passamonti, C., Frissen, I., & Làdavas, E. (2009). Visual recalibration of
auditory spatial perception: two separate neural circuits for perceptual learning.
European Journal of Neuroscience, 30, 1141-1150.
Rorden, C., & Karnath, H.-O. (2010). A simple measure of neglect severity.
Neuropsychologia, 48, 2758-2763.
Schuett, S., Heywood, C., Kentridge, R.W., & Zihl, J. (2008). The significance
of visual information processing in reading: insights from hemianopic dyslexia.
Neuropsychologia, 48, 2445-2462.
27
Siefer, A., Ehrenstein, W.H., Arnold-Schulz-Gahmen, B.E., Sökeland, J., &
Luttmann, A. (2003). Populationsstatistik und Assoziationsanalyse sensumotorischer
Seitenbevorzugung und deren Relevanz für verschiedene berufliche Tätigkeitsfelder.
Zentralblatt für Arbeitsmedizin, Arbeitsschutz und Ergonomie, 53, 346-353.
Strebel, J. (1924). Über Hemianopien. Archiv für Augenheilkunde, 94, 27-55.
Tham, K., & Tegnér, R. (1996). The baking tray task: a test of spatial neglect.
Neuropsychological Rehabilitation, 6, 19-25.
Weintraub, S., & Mesulam, M.-M. (1985). Mental state assessment of young
and elderly adults in behavioral neurology. In: M.-M. Mesulam (Ed.), Principles of
Behavioral Neurology (pp. 71–123). Philadelphia: Davis.
Zihl, J., & von Cramon, D. (1986). Zerebrale Sehstörungen. Stuttgart:
Kohlhammer.
Zimmer, U., Lewald, J., & Karnath, H.-O. (2003). Disturbed sound
lateralization in patients with spatial neglect. Journal of Cognitive Neuroscience, 15,
683-693.
28
Table 1. Summary of clinical data and visual field defects of patients with hemianopia.
Patien Ag Se
t
e
x
Side
VF
of HA border
Time since
onset
LHA1
38
F
L
-1.2°
7 months
LHA2
60
M
L
-2.9°
7 years
LHA3
64
M
L
-1.7°
LHA4
64
M
L
LHA5
42
F
LHA6
54
LHA7
Ethiology
Lesion site
AVM, ICH R temporo-parieto-occipital
ICH
R temporal
6 months
CI
R temporo-occipital
-4.6°
5 years
CI
R occipital
L
-0.3°
35 months
CI
R temporal
F
L
-0.2°
19 months
CI
R temporo-parieto-occipital
39
M
L
-1.8°
5 years
CI
R occipital
RHA1 44
M
R
12.5°
5 months
CI
L temporo-occipital
RHA2 48
M
R
0.4°
33 months
CI
L temporo-occipital
RHA3 37
F
R
8.2°
6 months
CI
L parieto-occipital, R occipital
Abbreviations AVM, cerebral arteriovenous malformation; CI, cerebral ischemia; F, female; ICH, intracerebral
hemorrhage; HA, hemianopia; L, left; M, male; R, right; VF, visual field. Negative angles are to the left, positive
angles to the right.
29
Figure 1. Results of Experiment 1 (visual pointing to acoustic targets). Final pointing
eccentricities (mean values ±SE) are plotted as a function of target azimuth for patients with
left (A) and right HA (B), and for matched controls (C). Data obtained for stimuli presented in
the left and right hemispaces were analysed separately. Responses were normalized such that
positive angles indicate pointing toward the hemispace within which the stimulus was
presented, and negative angles indicate pointing responses to the opposite hemispace.
Continuous lines indicate regression lines, dotted lines indicate ideal performance.
30
Figure 2. Results of Experiment 2 (acoustic pointing to visual targets). Final pointing
eccentricities (mean values ±SE) are plotted as a function of target azimuth for patients with
left (A) and right HA (B), and for matched controls (C). Conventions are as in Fig. 1.
31
Figure 3. Linear regression analysis of the pointing responses for Experiments 1 and
2. Each panel shows data (mean values ± SEM) obtained in the anopic and intact hemispaces
of patients with HA as well as data from matched controls. For the control subjects the
‘anopic hemispaces’ and ‘intact hemispaces’ were that hemispaces that were assigned as
controls to the anopic and intact hemispaces of patients with HA (see Materials and Methods).
(A and B) y-intercepts of the regression lines that were taken as a measure of constant error in
pointing. Data were normalized such that positive values indicate final deviations of the visual
stimulus from the auditory stimulus toward the anopic side, irrespective of whether the visual
stimulus was aligned with the auditory target (A) or the auditory stimulus was aligned with
the visual target (B). (C and D) Slopes of the regression lines that were taken as a measure of
the subject’s general tendency to underestimate (values < 1) or overestimate (values > 1)
distances between target positions. (E and F) Coefficients of determination of the regression
lines that were taken as a measure of precision in light pointing (E) and acoustic pointing (F).
32
Supplementary Material
Supplementary Table 1. Summary of lesion data. Areas involved by lesion are coded using the method of Damasio
and Damasio (1989)
Patient
Frontal
lobe
Temporal lobe
Parietal lobe Occipital lobe
LHA1
Right T6
Right P2
LHA2
Right T3, T4, T 6, T 9, T12
LHA3
Right T11
Right O1, O2, O3, O7
LHA4
Right T4
Right O1, O2, O4, O5
LHA5
Right T10, T11, T12
LHA6
Right F2
Right T4, T6
Right BG1
Right P2, P4 Right O1, O2, O3, O4, O5
Right O1, O2
Left F2
RHA2
RHA3
Right O4, O5
Right Th1, Th3, IC2
LHA7
RHA1
Central grey and
adjoining white
matter
Left O1, O2, O3, O6
Left T6, T10, T11
Left F2
Left O3, O6
Left P4
Left O1, O2, O3, O6
Right O4
33
Supplementary Figure 1. Visual field defects and lesion sites of patients with left
(LHA1-7) and right hemianopia (RHA1-3). (A) Reconstructions of the monocular central
visual fields based on static perimetry (up to 30° eccentricity; black areas, anopic regions;
white areas, intact regions). (B) Series of schematic brain slices along the superior-inferior
direction for each of the ten patients are depicted using standardized templates from Damasio
and Damasio (1989), with black areas indicating the lesioned sites. More inferior templates
are to left, more superior templates to the right. Templates are in neurological orientation, i.e.,
the left side of the template refers to the left side of the brain.
34
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement